MGCP gateways - Route List failover failure

Unanswered Question
Jul 18th, 2007
User Badges:

Hi people !


Please help me understand what's happening.


I have two 2801 with FXO ports, both of them registered with the CM and I can place calls throught them with no problem. They're MGCP gateways.


Now I've configured two Route Groups, one for each gateway and crated one Route List containing those two route groups, with different prefixes (because each gateways connects to different carriers).


So I place one call to the PSTN and it takes the first Route Group of the Route List, as expected. When I place the second call (simultaneously), it takes the second Route Group (also expected, because I have just one port per gateway). No problem here.


If I disconnect the phone line of the FXO of the primary Gateway , all outbound calls take the secondary route group (still fine), BUT when I re-connect the line of the primary gateway, the CallManager doesn't forward calls to it anymore ! All calls go through the secondary gateway, even if the primary FXO line is already connected and the voice port appear as Registered on the CallManager.


This condition is just reversed when I issue a "mgcp / no mgcp" command on the primary gateway. After this command it start to pass calls again as it is the first option according to the Route List configuration.


Can anyone tell me how can I solve this ?


Thanks and regards !!


Rodrigo

  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 0 (0 ratings)
Loading.
gogasca Wed, 07/18/2007 - 10:25
User Badges:
  • Green, 3000 points or more

Could you let us know the IOS version|CCM exact version.

Please gather CCM SDI|SDL traces and also debug mgcp packet and upload them.

We need to see if CCM is attempting to ruote the call to the GW and GW is not responsing 200 OK succesful, or CCM RL module is not recovering for that condition.

rcarrijoesilva Wed, 07/18/2007 - 11:42
User Badges:

Hi !!


IOS 12.(4)10 (image name: c2801-adventerprisek9-mz.124-10.bin)


CCM: 4.1(3)


Here's some outputs of "debug mgcp packets" when I tried to place a call during the problem:


*Jul 18 19:09:30.520: MGCP Packet received from 172.18.1.254:2427--->

200 330883180

<---


*Jul 18 19:09:32.272: MGCP Packet received from 172.18.1.254:2427--->

CRCX 587 [email protected]_1 MGCP 0.1

C: A000000001000106000000F5

X: 0

M: inactive

R: H/oc, H/of, L/hd

S: H/sup(addr(0,2,1,6,6,2,5,1,4))

Q: process,loop

<---


*Jul 18 19:09:32.272: MGCP Packet sent to 172.18.1.254:2427--->

400 587 Previous message in-progress

<---


*Jul 18 19:09:33.292: MGCP Packet received from 172.18.1.254:2427--->

CRCX 10587 [email protected]_1 MGCP 0.1

C: A000000001000106000000F5

X: 0

M: inactive

R: H/oc, H/of, L/hd

S: H/sup(addr(0,2,1,6,6,2,5,1,4))

Q: process,loop

<---


*Jul 18 19:09:33.296: MGCP Packet sent to 172.18.1.254:2427--->

400 10587 Previous message in-progress

<---


*Jul 18 19:09:33.296: MGCP Packet received from 172.18.1.254:2427--->

RQNT 588 [email protected]_1 MGCP 0.1

X: 0

R: L/hd

S:

Q: process,loop

<---


*Jul 18 19:09:33.296: MGCP Packet sent to 172.18.1.254:2427--->

400 588 Previous message in-progress

<---


*Jul 18 19:09:34.328: MGCP Packet received from 172.18.1.254:2427--->

RQNT 10588 [email protected]_1 MGCP 0.1

X: 0

R: L/hd

S:

Q: process,loop

<---


*Jul 18 19:09:34.328: MGCP Packet sent to 172.18.1.254:2427--->

400 10588 Previous message in-progress

<---


########################

Now the same debug after a "no mgcp / mgcp" commands:


*Jul 18 19:10:20.512: MGCP Packet received from 172.18.1.254:2427--->

CRCX 595 [email protected]_1 MGCP 0.1

C: A000000001000108000000F5

X: 0

M: inactive

R: H/oc, H/of, L/hd

S: H/sup(addr(0,2,1,6,6,2,5,1,4))

Q: process,loop

<---


*Jul 18 19:10:23.528: MGCP Packet received from 172.18.1.254:2427--->

CRCX 595 [email protected]_1 MGCP 0.1

C: A000000001000108000000F5

X: 0

M: inactive

R: H/oc, H/of, L/hd

S: H/sup(addr(0,2,1,6,6,2,5,1,4))

Q: process,loop

<---


*Jul 18 19:10:23.940: MGCP Packet sent to 172.18.1.254:2427--->

200 595 OK

I: 2


v=0

c=IN IP4 172.18.1.100

m=audio 18970 RTP/AVP 0 8 18 99 101 102 2 15 103 4 104 3 105 106 107 125 100

a=rtpmap:99 G.729a/8000

a=rtpmap:101 G.726-16/8000

a=rtpmap:102 G.726-24/8000

a=rtpmap:103 G.723.1-H/8000

a=rtpmap:104 G.723.1-L/8000

a=rtpmap:105 G.729b/8000

a=rtpmap:106 G.723.1a-H/8000

a=rtpmap:107 G.723.1a-L/8000

a=rtpmap:125 G.nX64/8000

a=rtpmap:100 X-NSE/8000

a=fmtp:100 192-194,200-202

a=X-sqn:0

a=X-cap: 1 audio RTP/AVP 100

a=X-cpar: a=rtpmap:100 X-NSE/8000

a=X-cpar: a=fmtp:100 192-194,200-202

a=X-cap: 2 image udptl t38

<---


*Jul 18 19:10:23.940: MGCP Packet sent to 172.18.1.254:2427--->

NTFY 330883188 [email protected]_1 MGCP 0.1

N: [email protected]:2427

X: 0

O: L/hd

<---


########################


I also noted something that may help: when I disconnect the fxo line from the primary gateway and try to make a call, I see (debug voice ccapi inout) that some messages come to the gateway and, as there's no line, it alerts the CCM, then the CCM redirects the call to the secondary gateway. After this redirecting, I see that the gateway downloads a new .xml configuration file. Somehow, this same gateway will accept new calls only after the "no mgcp/mgcp" when a new .xml comes from the CCM.

I removed the "ccm-manager config" command and it seems that the router didn't received/asked for a new xml file, but it didn't help it anyway; the problem is still happening...


I'll try to collect SDL files...


Thanks for your attention !

Regards,

Rodrigo

rcarrijoesilva Wed, 07/18/2007 - 12:04
User Badges:

Hello


Attached follows a SDI Trace. I guess that the problem is on the router...

The failed gateway is 172.18.1.100.

The gateway that accept the call is 172.18.1.101.


Thanks !

Rodrigo

rcarrijoesilva Wed, 07/18/2007 - 12:52
User Badges:

Hi !


I've uploaded the gateway to 12.4(3g) and everything is working fine now !

Thanks for you time !


Regards,

Rodrigo

Actions

This Discussion